Show HN: A free, privacy preserving, archive of public Discord servers

Okay, the "not really" and "I'll solve that problem if and when" responses are... something else. It feels like you're speedrunning how to get into a world of trouble while hand-waving away every legitimate concern. Let's try to unpack this again, because your justifications are frankly baffling.

> Again, not really. It's impossible to search for users without already knowing what server they are in. This is functionally identical to Discord's in-built search feature.

That's not quite correct, and frankly it borders on willful obfuscation. In your own words elsewhere in this thread, you're eager for search engines to index this archive. That "privacy preserving" barrier of needing to know both a user ID and a server/channel id evaporates the moment Google or any other search engine hoovers up your pages. At that point, any combination of keywords, usernames, aliases, or snippets could reveal someone's posting history, across contexts and years. How is that "functionally identical" to Discord's walled-garden search or "privacy preserving"?

> I believe that people need to realize that their messages were already being logged by many different moderation bots, just not publicized.

This is a disingenuous deflection.

  - Moderation bots operate within a specific server, for a specific purpose (moderation, utility) defined by the server admins. Their logs are typically for admin/moderator use, not for creating a global, publicly searchable archive.
  - Users joining a server often see these bots, understand their function, and server admins explicitly add these bots. It's a known quantity. What you're doing is orders of magnitude different - an external, uninvited entity scraping everything discoverable and making it universally public.
  - Just a matter of time" is a lazy, fatalistic excuse for unethical data harvesting. Just because something can be technically scraped doesn't mean it should be, or that you doing so is fine.
    

Your "I really do want to build a platform that balances privacy and usability" line sounds utterly hollow when the entire foundation of the project demonstrates a profound misunderstanding, or disregard, for basic privacy, consent, and intellectual property.

Speaking of which... have you actually thought about the legal Pandora's Box you're prying open? Your casual "I'll deal with Discord's ToS issues if they arise" attitude is quaint, because Discord's ToS is likely the tip of a colossal iceberg of legal trouble.

You're not just 'breaking ToS', you're potentially looking at:

  - Data Protection Law Violations (GDPR, CCPA, etc.) because you're scraping personal data of EU/California (and other) residents without any lawful basis. The fines can be astronomical. "Opt-out" after the fact for data you had no right to take in the first place isn't how this works.
  - COPPA Violations if you scraped any messages from a 12-year-old on a "public, discoverable" server before their account was deleted by Discord. Guess who's holding that data now without parental consent? You.
  - Every original, creative message is copyrighted by its author. Roleplay, detailed discussions, code snippets, even well-crafted tirades – you're republishing millions of these. While not every "lol" is copyrightable, a massive volume of content on Discord absolutely is. "Fair use" for wholesale, non-transformative republication on this scale? Unlikely.
  - And last but not least, CSAM (Child Sexual Abuse Material): This is the nightmare scenario. You are scraping public Discord. Some public, poorly moderated Discords inevitably contain links to or text-based CSAM. Even if you don't intend to host it, if your scraper picks it up and it becomes accessible via your archive (even just a link), you are in profoundly serious trouble. "But I don't re-publish attachments" is irrelevant if you're archiving and re-publishing the links. This isn't just fines; this is potential prison time.

Good luck with all of this.

I hope you have a good lawyer, ideally multiple. You might need them.

How does this tool handle it when a user deletes or edits their message? Is the message still logged on your site?

I was imagining something like a month-by-month historical export so only recent data would need to be regularly exported.

SQLite dumps would be awesome to work with though.

No, what's "ridiculous" is this simplistic, black-and-white framing that deliberately ignores any nuance, the concept of contextual integrity or reasonable user expectations.

Of course, no one expects absolute secrecy in a public-facing Discord server. That's a straw man. The issue isn't about some naive belief that messages are invisible. It's about the scope, permanence, and method of access and archiving.

People participating in public Discord spaces have reasonable contextual expectations about how their words will be accessed and by whom. They expect their messages to be seen by current and maybe future server members - not extracted, permanently archived, and made globally searchable by entirely unrelated third parties.

This is similar to how conversations in a public park are technically "public," but most people would be rightfully disturbed if someone recorded everything, transcribed it, published it online with their names attached, and made it all searchable forever. Just because something isn't strictly private doesn't mean any and all forms of collection, republication, and indexing are ethically justified.

If you can't see the distinction between "not perfectly private within this specific semi-public space" and "archived indefinitely, and globally searchable forever by anyone, anywhere, for any reason," then you're either arguing in bad faith or your understanding of these issues is so superficial that further engagement is pointless.